Presentation at multi-paper session, Scaling Up and Scaling Down: Interrogating Data to Find Widespread Trends While Honoring Local Contexts, at the 2019 American Evaluation Association [AEA] 33rd Annual Conference

Time: 11:30 a.m.–12:15 p.m.

Location: Minneapolis Convention Center, Hilton Minneapolis, Minnesota

Description: This presentation will examine the role of biases and downfalls in using linear regression analyses and will present alternate analysis methods used in research that may allow for emergent patterns and categories from the data. The presentation will include an overview of multiple machine-learning analytic models, including classification and regression trees (CART) and cluster analysis, that can be applied to student college placements in the Pacific Region and other contexts in which researchers need to examine emergent data structures that may be invisible with traditional analyses.

The presentation will be grounded in REL Pacific’s contextual approach to conducting research in the Pacific Region, included current proposed studies on Using High School and College Data to Predict Teacher Candidates’ Performance on the Praxis Core® Exam at the Unibetsedåt Guåhan (University of Guam) and Factors Associated with Third Grade Reading Outcomes of Students in the Commonwealth of Northern Marianas Islands Public School System), and is intended to share information about methodologies and reporting structures that are easily accessible to practitioners and that may foster their understanding of findings within diverse contexts and in turn facilitate changes in practice.

Audience: The primary audience for this event will be participants at the 2019 AEA Conference in Minneapolis, Minnesota.

Speakers: Dr. Bradley Rentz, REL Pacific at McREL International
Dr. Sheila A. Arens, REL Pacific at McREL International